can-i-run-this-llm

Can I run this LLM?

Can a AMD Instinct MI300X (192GB) run Gemma 4 26B-A4B (MoE)?

Gemma · Apache 2.0 · estimated for a 192.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Gemma 4 26B-A4B (MoE) at FP16 · 50.5 GB · 206–310 tok/s on a AMD Instinct MI300X (192GB).

Every quantization on this machine

QuantSizeSpeedFit
Q3_K_M12.7 GB302–452 tok/sRuns great
Q4_K_M17.0 GB287–430 tok/sRuns great
Q5_K_M21.2 GB274–410 tok/sRuns great
Q6_K23.2 GB268–401 tok/sRuns great
Q8_026.9 GB257–386 tok/sRuns great
FP1650.5 GB206–310 tok/sRuns great

Ran Gemma 4 26B-A4B (MoE) on a AMD Instinct MI300X (192GB)? Add your real numbers →